Downloaded from orbit.dtu.dk on: Sep 24, 2021

Do pout ( esmarkii) die from spawning stress? Mortality of Norway pout in relation to growth, maturity and density in the North Sea, Skagerrak and Kattegat Nielsen, J. Rasmus; Lambert, G.; Bastardie, Francois; Sparholt, H.; Vinther, Morten

Published in: ICES Journal of Marine Science

Link to article, DOI: 10.1093/icesjms/fss001

Publication date: 2012

Document Version Publisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA): Nielsen, J. R., Lambert, G., Bastardie, F., Sparholt, H., & Vinther, M. (2012). Do Norway pout (Trisopterus esmarkii) die from spawning stress? Mortality of Norway pout in relation to growth, maturity and density in the North Sea, Skagerrak and Kattegat. ICES Journal of Marine Science, 69(2), 197-207. https://doi.org/10.1093/icesjms/fss001

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.  You may not further distribute the material or use it for any profit-making activity or commercial gain  You may freely distribute the URL identifying the publication in the public portal

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

ICES Journal of Marine Science

ICES Journal of Marine Science (2012), 69(2), 197–207. doi:10.1093/icesjms/fss001

Do Norway pout (Trisopterus esmarkii) die from spawning stress? Mortality of Norway pout in relation to growth, sexual maturity, and density in the North Sea, Skagerrak, and Kattegat

J. Rasmus Nielsen1*‡, Gwladys Lambert1‡, Francois Bastardie1, Henrik Sparholt2, and Morten Vinther1 1National Institute of Aquatic Resources (DTU-Aqua), Technical University of Denmark, Charlottenlund Castle, DK-2920 Charlottenlund, Denmark Downloaded from 2ICES, H. C. Andersens-Boulevard 44-46, DK-1553 Copenhagen V, Denmark *Corresponding author: tel: +45 35 88 33 81; fax: +45 35 88 33 33; e-mail: [email protected]

Nielsen, J. R., Lambert, G., Bastardie, F., Sparholt, H., and Vinther, M. 2012. Do Norway pout (Trisopterus esmarkii) die from spawning stress?

Mortality of Norway pout in relation to growth, sexual maturity, and density in the North Sea, Skagerrak, and Kattegat. – ICES Journal of http://icesjms.oxfordjournals.org/ Marine Science, 69: 197–207. Received 25 February 2011; accepted 29 December 2011.

The mortality patterns of Norway pout (NP) are not well understood. It has been suggested that NP undergo heavy spawning mor- tality, and this paper summarizes and provides new evidence in support of this hypothesis. The very low–absent fishing activity in recent years provides a unique opportunity to analyse the natural life-history traits of cohorts in the NP stock in the North Sea. Based on the ICES trawl survey abundance indices, cohort mortality is found to significantly increase with age. We argue that this cannot be explained by selectiveness in the fishery, potential size-specific migrations out of the area, higher predation pressure on older individuals, or differences in survey catchability by NP age from before to after spawning and that it is higher in the main spawn- at DTU Library on February 21, 2012 ing areas than outside. We found that natural mortality (M) is significantly correlated with sexual maturity, sex, growth, and intra- specific stock density. All of this is consistent with a greater mortality occurring mainly from the first to the second quarter of the year, i.e. spawning mortality, which is discussed as being a major direct and indirect cause of stock mortality. Keywords: cohort analysis, density-dependence, growth, maturity, natural and fishing mortality, North Sea, Norway pout, population dynamics, spawning, spawning stress and mortality, Trisopterus esmarkii.

Introduction analysis) model, and the SURBA (survey-based assessment) The North Sea–Skagerrak–Kattegat Norway pout (NP; model (ICES, 2004, 2006, 2008; Supplementary material). Trisopterus esmarkii) stock is an important food source for com- Despite these differences, constant values of natural mortality of mercially important fish species, such as cod (Gadus morhua), M ¼ 0.4 per quarter for all ages are still used in the ICES single- saithe (Pollachius virens), (Melanogrammus aeglefinus), stock analytical assessment (ICES, 2010). mackerel (Scomber scombrus), and whiting (Merlangius merlan- Although mortality by predation of the NP stock decreases or gus). Therefore, this small, short-lived species is an important remains somewhat constant as fish grow older, based on docu- prey organism in the North Sea ecosystem (Sparholt et al., mentation from existing stomach sampling programmes and 2002a; ICES, 2008; Rindorf et al., 2010). In addition, the NP MSVPA analyses (Sparholt, 1994; ICES, 2006, 2008; Rindorf stock is usually a direct target of a significant small-meshed et al., 2010), total natural mortality increases with age (Sparholt fishery for reduction (industrial) purposes (ICES, 2007a, b, c, et al., 2002a, b). Total mortality is also substantially higher than 2010). the fishing mortality documented through the ICES single-stock The time-series of the NP stock mortality shows substantial dif- assessments (Sparholt et al., 2002a; ICES 2007a, b, c, 2008, ferences between natural mortality by age as estimated by Sparholt 2010). As a result, total mortality (Z) cannot be exclusively et al. (2002a), the MSVPA (multispecies virtual population explained by fishing activities and direct predation mortality;

‡Equal authorship.

# 2012 International Council for the Exploration of the Sea. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 198 J. R. Nielsen et al. there is another important source of natural mortality. The highest small gadoid species dies abruptly and at a relatively young age total mortality rates have been observed between the first (Q1) and from spawning stress (e.g. Ursin, 1963; Bailey and Kunzlik, 1984; second (Q2) quarters of the year, which correspond to the spawn- Lambert et al., 2009) and energy depletion, similar to some ing season (Sparholt et al., 2002b; Lambert et al., 2009). salmon species, capelin, anchovies, and gobies, or if there are Residual mortality, i.e. natural mortality caused by factors other other reasons for which they have a short lifespan. In the study than predation, is not well known or documented for fish in by Lambert et al. (2009), maturity and growth dynamics were general nor are the processes contributing to it (Baur et al., thoroughly investigated. In the present paper, we analyse the 2006; Bass et al., 2007; Golubev, 2009; Gislason et al., 2010; level of natural mortality in relation to maturity, sex, and growth Partridge, 2010). Several small, short-lived fish species have dynamics on quarterly and geographically disaggregated bases. adult natural mortalities of more than 0.6 that increase with age This is done by the use of long-term data time-series in an effort (e.g. Gislason et al., 2010). The increase in M with age during to understand the mechanisms behind the dynamics of mortality, the adult life stage is, however, not well investigated for fish including predation and fishing mortality. Lambert et al. (2009) because it is difficult to isolate M from fishing mortality (F), but and the present study have different objectives, but overlapping it has been documented for some stocks of small fish species documentation, and some figures from Lambert et al. (2009) (Beverton, 1963; Caputo et al., 2002; Cook, 2004; Terzibasi et al., have, therefore, been used in the present study. We tested three 2007; Golubev, 2009; Uriarte et al., 2010). For several fish null hypotheses: (i) H01: natural mortality is constant over years

species, there is evidence of residual mortality as a result of (at a level of approximately M ¼ 1.6) and quarters (M ¼ 0.4) Downloaded from active gene-directed and age-determined apoptosis, senescence, and independent of age; (ii) H02: there is no relationship and diseases associated with spawning. For some short-lived fish between natural mortality and reproduction-specific life-history species, this is associated with truncated ontogeny, accelerated traits of NP, such as sex, maturity, or growth and, thus, mortality gonad maturation, and spawning events (e.g. Caputo et al., is decoupled from spawning; and (iii) H03: there is no density- 2002; Terzibasi et al., 2007). Mediterranean goby (Aphia minuta) dependence, neither intra- nor interspecific, in NP mortality. seems to have an endogeneous timer-inducing adult mortality im- http://icesjms.oxfordjournals.org/ mediately after the first spawning season by causing irreversible in- Material and methods testinal deterioration (Caputo et al., 2002). Age-dependent To complete the objectives of this study, extensive disaggregated degeneration or the dysfunction of several organs and age-related data were used, involving complex data compilation, manipula- pathological changes similar to those of mammals has been tion, and analyses (see also Lambert et al., 2009). The yearly abun- demonstrated for a variety of fish species (Woodhead, 1998; dance indices were computed from survey raw catch per unit effort Kishi et al., 2003; Reznick et al., 2006; Buston and Garcia, 2007). (cpue) data by fish length combined with raw sex–maturity age– The short lifespan of the fish Nothobranchius furzeri is associated length keys (SMALKs). Data were available from the ICES coordi- with explosive growth, accelerated sexual maturation, and the ex- nated International Bottom Trawl Survey (IBTS) for 1983–2006 pression of ageing-genes causing behavioural and histological covering the North Sea and Skagerrak–Kattegat (Anon., 2004). at DTU Library on February 21, 2012 changes (Terzibasi et al., 2007). Spawning mortality is observed These indices were stratified by roundfish areas (RFAs) 1, 2, 3, 4, for other small, short-lived fish species such as capelin (Mallotus 7, 8, and 9 (see RFAs in Figure 1 in Lambert et al., 2009). The villosus) and anchovy (Engraulis encrasicolus; Uriarte et al., 2010) total area differs and is wider than the combined index area and suggested for Northeast Arctic cod (G. morhua), where covered for the standard calculation of the ICES IBTS abundance males mature earlier and have higher mature mortality than indices to assure the coverage of the full NP stock distribution area females (Jakobsen and Ajiad, 1999). Indirect spawning-related needed for the area-disaggregated analyses of geographical vari- mortality may also originate from the abrupt and substantial ability (Anon., 2001, 2004; Sparholt et al., 2002a, b; ICES, 2004, energy loss with increased vulnerability and exposure to inverte- 2007a, b, c). The NP stock is distributed mainly in the northern brate scavengers and predators. However, to our knowledge, it North Sea, Skagerrak, and Kattegat, and potential seasonal migra- has not been reported for small species of the family. tions do not result in the migration out of this shelf area because In the North Sea, small gadoids, such as blue whiting the NP stock is not distributed in areas with depths greater than (Micromesistius poutassou), poor cod (Trisopterus minutus), 200–250 m (Sparholt et al., 2002a, b). Data manipulation was ne- pouting (Trisopterus luscus), silvery cod (Gadiculus argenteus), cessary to perform the analyses of geographical variation because and NP, have been observed living up to ages 10–20, 5, 4, 3, the SMALKs were not always complete for all areas for every and 5 years, respectively (www.fishbase.org), indicating that year and quarter (i.e. for each cell). The initial data were taken spawning stress and mortality might be an issue for some of from the ICES DATRAS database, and where information was these species. missing, empty cells were filled with estimates based on the Previously, data had been inadequate to investigate these dy- methods given in the IBTS manual and on biological and ecologic- namics for NP; however, there is a unique opportunity to estimate al knowledge (Anon., 2001, 2004; ICES 2007d). Such data filling M for this normally exploited stock because fishing activity has was necessary in the SMALKs because there was a risk to bias been very low since 2003. The targeted fishery was totally closed data and results if valid and available cpue information were in 2005 and 2007 as well as for the first half of 2006 because of a excluded from cells and not used in the analyses because there low stock level. In those periods, M approximately equals total were no SMALK observations for these cells (Hoenig and mortality (Z). Heisey, 1987). The manual provides standard substitution proce- It is essential, for both ecosystem and single-stock manage- dures for converting length data to age data when age–length ment, to investigate the natural mortality dynamics of NP, the keys (ALKs) are missing (for one RFA, 1 year, and one quarter) periodical variability herein, and to provide accurate data on life- or are not reliable (e.g. an RFA with , 25 otoliths sampled). history traits influencing mortality rates used in ICES analytical Following the IBTS manual, the ALKs of certain RFA were used assessments. Therefore, it is important to know whether this to replace the missing observations of the neighbouring RFAs in Do Norway pout (Trisopterus esmarkii) die from spawning stress? 199 a given quarter and year, when needed. However, the same proced- The disaggregated Z-values were, in a few cases, estimated to be ure for converting data to sex and maturity data used in Lambert negative, particularly from Q1 to Q2 and from Q3 to Q4, which is et al. (2009) was not adequate for the present study because of the likely a consequence of incomplete spatial coverage in Q2 and Q4. spawning migration out of Skagerrak–Kattegat and its related sex ICES has evaluated the quality of the IBTS Q1 and Q3 to be high, distribution patterns (Ursin, 1963; Poulsen, 1968; Lambert et al., and those quarters are estimated by ICES to be consistent with 2009). Maturity ratios have proven variable between years respect to coverage and catchability and are used in stock assess- (Lambert et al., 2009), whereas sex patterns are more consistent. ments (Fraser et al., 2007; ICES, 2007a, b, c). Consequently, the Consequently, the missing observations in the SALKs were indices from Q1 and Q3 were mainly used in this study. replaced by the average of all available years for the same quarter For H01, the IBTS estimates of Z by age were compared with the and RFA. Therefore, total abundance indices, both in total and MSVPA and SURBA model estimates (ICES, 2004, 2006, 2008) by sex, were computed following the procedure described by and to the estimates from Sparholt et al. (2002a) to evaluate ICES (Anon., 2001). seasonal and long-term trends in mortality as well as its Total mortality (Z) was calculated from the cpue values age-dependence. Emphasis was placed on the more recent (Ricker, 1975; Sparre and Venema, 1989): period when fishing activity, and fishing mortality were very low or zero in the NP fishery. 1 cpue(t ) For H02, the evidence for linking mortality patterns to maturity Z = ln 1 , (1) and growth dynamics, i.e. indicating potential spawning mortality, Downloaded from t2 − t1 cpue(t2) was summarized based on Lambert et al. (2009). The mortality of mature individuals could not be computed directly because the where cpue is the catch in the number of individuals per trawl percentage of fish maturing from one quarter to the next (from, hour, and t and t represent ages, with t , t . 1 2 1 2 e.g. histological studies) is unknown (Lambert et al., 2009). To perform a robustness check and sensitivity analysis on the Therefore, alternative multiple linear regressions and analyses of above described data-manipulation procedures for filling in gaps http://icesjms.oxfordjournals.org/ variances were performed to check for consistency between total in SMALKs and SALKs, the yearly Z values by age were computed mortality (Z) and sex and maturity ratios. with Equation (1) using the revised area-disaggregated IBTS cpue Concerning H 3, variation in growth and maturity has been data as described above (for age groups 1–4+) and compared 0 shown to be dependent on both intra- and interspecific densities with mortality estimates from abundance indices using the ICES (Lambert et al., 2009). The mortality rates were consequently ana- standard calculation procedures and area. The comparison lysed in the context of variations in density. Linear regressions of Z showed that the dynamics of the mortality from the revised cpue as a function of the NP stock numbers and biomasses were tested indices were very similar to those of ICES (Table 1). The present and were also tested as a function of the spawning-stock biomasses data compilation establishing more disaggregated data was thus (SSBs) of the main predator stocks of cod, saithe, haddock, determined to be valid and was preferred to investigate quarterly mackerel, and whiting (ICES, 2007a, 2009). and sexual-disaggregated mortality patterns. at DTU Library on February 21, 2012 Results Table 1. Total mortality (Z) calculated based on IBTS cpue data Magnitude and variability of mortality by age (H01) according to ICES standard calculation procedures and according Z by age does not show periodical trends over the period 1983– to the revised calculation procedure. 2006, except seasonal trends, and Z increases with age for all

Cohort Z1–2 ICES Z1–2 (revised) Z2–3 ICES Z2–3 (revised) years (Figures 1 and 2; Supplementary material), i.e. both in 1981 – – 2.07 2.52 years with and without targeted fishery, so the fishery cannot 1982 0.83 0.84 2.60 2.56 explain the difference. This age-dependent mortality is consistent 1983 1.25 1.23 4.27 4.08 for Zage3, but the few exceptions here may be artefacts because of 1984 1.81 1.74 1.84 1.91 the scarcity of age group 4 individuals observed. For 2005, Z cor- 1985 1.46 1.37 3.47 3.56 responds to M, because the fishery was closed in that year 1986 1.48 1.38 1.43 1.58 (Figure 1). Here, Mage1 (1.59) is equivalent to the value used in 1987 20.72 20.55 1.88 1.89 the assessment, Z ¼ 1.6. M and M are higher at 1.81 and 1988 0.99 1.03 1.75 1.35 age2 age3 2.11, respectively. This age-dependent mortality confirms the con- 1989 0.60 0.52 3.10 3.14 1990 1.02 0.96 1.23 1.26 clusions obtained from the SURBA model analyses performed by 1991 0.65 0.67 3.69 3.60 ICES (ICES, 2004; Supplementary material) and from Sparholt 1992 1.97 1.89 1.58 1.53 et al. (2002a), while contradicting the MSVPA outputs indicating 1993 0.85 0.85 1.24 1.31 constant natural mortalities by age (ICES, 2004, 2006, 2007a, 1994 0.81 0.84 1.37 1.44 2008). Both the 2008 and the ICES WGSAM 2011 results show 1995 20.47 0.26 1.72 1.75 some annual variability in the rate of mortality by predation 1996 0.60 0.66 2.08 2.08 (M2), but a similar level for M2 at ages 1 and 2. 1997 0.53 0.60 2.22 2.14 1998 0.83 0.71 1.88 1.91 1999 1.04 1.02 1.18 1.18 Dynamics of maturity, spawning time, and place 2000 0.48 0.64 2.15 2.10 in relation to mortality (H02) 2001 1.14 1.00 2.83 2.81 From Lambert et al. (2009), we know that the ratio of mature indi- 2002 1.19 0.96 2.32 2.47 viduals at ages 2 and 3 decreases from Q1, i.e. spawning time, to 2003 1.92 1.79 1.58 1.81 Q3 (Figures 2 and 13 in Lambert et al., 2009). In addition, very 2004 1.55 1.59 – – few post-spawning fish have been recorded despite extensive 200 J. R. Nielsen et al. Downloaded from

Figure 1. Total mortality (Z) by age over a 23-year period calculated according to Equation (1) based on revised IBTS Q1 cpue data. The negative value from 1988 age 1 was omitted from the calculation. http://icesjms.oxfordjournals.org/ at DTU Library on February 21, 2012

Figure 2. Seasonal total mortalities (Z) by sex and age for strong and weak year classes based on revised IBTS Q1 and Q3 cpue data. Z is calculated according to Equation (1). Error bars represent the standard deviations.

survey efforts, indicating a high mortality of mature individuals Figure 3. Total mortality (Z) of females (black dots) and males following the spawning event (Figure 2 in Lambert et al., 2009). (white dots) as a function of the fraction mature for age groups 1 The maturity ratio also increases with age for both sexes and and 2. Z is calculated according to Equation (1) and based on the shows a strong spatial pattern, reflecting likely effects of spawning, revised IBTS cpue data. Regression t-test statistics: p , 0.001 for as explained below (Figures 2, 3, 13, and 14 and Table 3 in Lambert females and p ¼ 0.058 for males. et al., 2009). Spawning areas of the NP stock are identified in Lambert et al. (2009), and the percentage of mature individuals is significantly higher in the main spawning areas RFA1 and are mature, and a very high level above 90% are mature. The RFA3, where the decrease in the maturity ratio from Q1 to Q3 is spread around the latter high level indicates that some other most evident (Figures 3 and 13 and Table 3 in Lambert et al., factors apart from the fraction mature potentially influence 2009). This strongly indicates a link between spawning and mortality. greater mortality during the breeding season, i.e. direct or indirect mortality caused by spawning stress. Although mortality cannot be Growth dynamics in relation to mortality (H02) directly calculated for the spawning areas during and just after the The growth of NP shows strong spatio–temporal differences spawning period, the yearly total mortality for both sexes is posi- (Figure 8 and Table 5 in Lambert et al., 2009). Body weight is gen- tively correlated with the overall maturity ratio assessed during the erally stable from Q1 to Q2, with a notable exception found in the spawning season (Figure 3 and Table 2; Figures 2, 3, and 13–15 western North Sea, where age group 2 loses considerable weight. and Table 3 in Lambert et al., 2009). Figure 3 shows a low mortality This is likely to be spawning-related because this area covers the rate by sex until 50% of the individuals are mature (with an un- main spawning ground of the stock. Besides the loss in weight, a explained gap between 0.4 and 0.6), a much higher rate above 60% general decline in mean length-at-age from Q1 to Q2 is also Do Norway pout (Trisopterus esmarkii) die from spawning stress? 201

Table 2. Statistics: F-test and the corresponding p-values of the to be statistically significant (Figure 2). The negative relationship multiple linear regression of total mortality (Z) from Q1 vs. sex and between year-class strength and the mortality rate of young age fraction mature. groups was investigated further. It appears that total mortality of Estimate s.d. t-value p(>|t|) age 1 males and females (Figure 6) tends to be lower when density is higher, although not significantly. No intraspecific Intercept 0.69 0.16 4.38 ,0.001 Sex: female 20.29 0.15 21.92 0.06 relationship between mortality and density was observed for Fraction mature 1.63 0.21 7.61 ,0.001 age 2 or 3. The decreasing pattern in mortality in relation to intraspecific density over time (Figure 6) is unlikely to be caused by predation, even if it is generally accepted that higher prey density usually results in overall lower predation mortality (ICES, 2006, 2008). No significant interspecific density-dependence in Z for age 1 or 2 was found in relation to SSB for the most important predator stocks in the North Sea known to prey upon NP (ICES, 2006, 2008), i.e. saithe, haddock, cod, and whiting (Figure 7), nor has this been found for Z by sex (not shown). A general pattern indi-

cates that mortality tends to increase when the main predator Downloaded from stocks become more abundant (Figure 7). However, this is not sig- nificant, and no seasonal patterns have been found to explain the quarterly patterns observed in the increasing mortality by age (Figure 2). Lambert et al. (2009) showed that the maturity rates of age 1 NP were also negatively influenced by density (Figures 5 and 15 http://icesjms.oxfordjournals.org/ in Lambert et al., 2009). The age and the length at which 50% of the fish are mature increase with increasing recruitment (Figure 18 in Lambert et al., 2009). Variations in natural mortality can consequently be explained by spawning mortality, i.e. maturity occurs later and, thus, spawning mortality is lower at high population densities. This correlation is supported by the relationships between Figure 4. Relationship between the fraction mature and the mean growth, density, and mortality. The mean length at age 1 length-at-age MLA (A1 Q1; males, p , 0.001; females, p , 0.001). (MLA1) Q1 is negatively correlated with density (Figure 8), and at DTU Library on February 21, 2012 Females, white dots and continuous curves; males, black dots and MLA1 Q1 is lower for strong year classes (female, p ¼ 0.05; dashed curves (from Lambert et al., 2009). male, p ¼ 0.03). At age 2, the decrease is only significant for females (p ¼ 0.04). observed for both females and males (Figure 8 and Table 5 in Lambert et al., 2009), implying that the proportion of large Discussion individuals has decreased from before to after spawning. H01: natural mortality is constant over years and quarters and independent of age Integrated growth and maturity patterns in relation This hypothesis is rejected. The present study shows that the to mortality (H02) annual NP total mortality Z varied over the last 25 years, but no When recruits benefit from favourable growth conditions during overall periodical trend could be observed except seasonal vari- their first year, i.e. reach relatively large mean length at age 1 in ation. There is a distinct and consistent age difference in Z. Z the first quarter (MLA1 Q1), more individuals will mature increases significantly with age and, based on Sparholt et al. before the spawning season (Figure 4). This clear pattern is (2002a), the peak in the length distribution representing larger observed for both females and males. Males mature earlier and individuals disappears between Q1 and Q2. The total mortality at smaller sizes than females (Figure 5). Consequently, males from 2005 to 2006 shows the same age pattern as for the full form the major part of the spawning stock during the first spawn- period investigated in the present study, i.e. mortality increases ing season of a cohort. This pattern coincides with the age 1 and 2 from ages 1 to 3. For these years, mortality corresponds to the mortalities for males being higher than for females in Q1 actual natural mortality because the fishery was closed at that (Figure 2), where males undergo a significantly greater mortality time. Both in historical times of relatively higher fishing mortality than females by an average of 0.2 (paired t-test, t ¼ 3.059, d.f. ¼ and in the most recent years of low fishing intensity, total mortality 101, p ¼ 0.003), potentially explained by spawning-associated was highest for the oldest fish. Consequently, the effect of higher mortality. fishing activity on the oldest age groups cannot explain the observed trend. Furthermore, fishing intensity and mortality Density-dependence in mortality related to (ICES, 2010) in the directed NP fishery are actually highest in density-dependence in maturity and growth (H03) Q3 and Q4, which cannot explain the higher Z in Q1 and Q2 Early ages of less abundant cohorts show consistently higher mean (Lambert et al., 2009), i.e. the seasonal patterns in Z. Also, no sex- mortality rates than the more abundant cohorts of both sexes, selective fishery was evident from the biological sampling from the although the standard deviations are too high for this difference fishery (not shown) that would explain the sexual differences 202 J. R. Nielsen et al. Downloaded from http://icesjms.oxfordjournals.org/

Figure 5. Fraction mature as functions of age [logit(p) ¼ a + b × age] (left) and length [logit(p) ¼ a + b × length] (right). Females, continuous lines; males, dashed lines; LC, length class. Vertical lines represent the age at 25, 50, and 75% maturity (from Lambert et al., 2009). observed in Z. Estimates of Z at age with confidence limits from a software R) based on IBTS Q1 and Q3 NP cpue data at age confirm stock assessment with the full population dynamic SURBAR the increasing mortality with increasing age (Supplementary model (SURBA standard ICES assessment model in the statistical Figure S1). Bootstrap analyses of observation variability (CV)in at DTU Library on February 21, 2012

2 = 2 = Figure 6. Total mortality (Z) based on revised IBTS Q1 cpue at age 1 vs. NP age 1 stock number (SN; rf 0.08, p ¼ 0.222; rm 0.14, p ¼ 2 = 2 = 2 = 2 = 0.106), spawning-stock number (SSN; rf 0.11, p ¼ 0.145; rm 0.10, p ¼ 0.178), SSB (t; rf 0.00, p ¼ 0.807; rm 0.00, p ¼ 0.942), total 2 = 2 = 2 = 2 = stock number (TSN; rf 0.09, p ¼ 0.177; rm 0.15, p ¼ 0.096), and total-stock biomass (TSB; t) (rf 0.12, p ¼ 0.117; rm 0.15, p ¼ 0.089). Female figures at left, and male figures at right; regression lines are shown; numbers in millions and biomass in tonnes (t). Z is calculated according to Equation (1). Do Norway pout (Trisopterus esmarkii) die from spawning stress? 203 Downloaded from http://icesjms.oxfordjournals.org/

Figure 7. Total mortality (Z) based on revised IBTS Q1 cpue at age 1 (top panels) and age 2 (bottom panels) vs. SSBs (t) of three main predators on 1 January. Regression lines of the relationships shown for cod (Cod; age 1, r2 ¼ 0; age 2, r2 ¼ 0.08), saithe (Sai; age 1, r2 ¼ 0.04; age 2, r2 ¼ 0), and haddock (Had; age 1, r2 ¼ 0.06; age 2, r2 ¼ 0.03). Z is calculated according to Equation (1). at DTU Library on February 21, 2012

the same data support the differences in Z at age not being just random variability (Supplementary Table S1). In brief, the observed significant age- and season-specific mortality patterns can be directly explained by greater mortality associated with the spawning event for older fish in the first part of the year.

Survey coverage and catchability with respect to the hypotheses H01–H03 A potential problem resulting from the use of survey results is that the sample size is generally small, and hence the abundance esti- mates are likely to be noisy (Cook, 1997). However, the survey time-series used in the present study have extensive coverage, and enough individuals have been sampled to obtain statistically significant results. ICES has evaluated the IBTS Q1 and Q3 and has concluded that they have adequate coverage and consistent time-series information for use in NP stock assessments (ICES, 2007a, b, c, d). These data are widely used in similar fish popula- tion dynamic analyses on NP and other demersal, gadoid species (e.g. Cook, 1997; Cotter, 2001; Beare et al., 2002; Lambert et al., 2009). The robustness and sensitivity analysis of potential data compilation effects of SMALKs to include all available cpue data Figure 8. Mean length-at-age in Q1 of age 1 (left) and of age 2 in Q1 for a wider area showed similar dynamics of the mortalities calcu- (right) vs. year-class strength [recruitment (R) of a cohort] showing lated from the revised cpue indices compared with those from the statistically significant intraspecific density-dependence. Females, ICES standard area. In both cases, the ICES standard calculation white circles and continuous lines; males, black dots and dashed (summing and raising) procedures were used (Anon., 2001). lines; cohorts in millions (from Lambert et al., 2009). The data filling has not been so extensive that it can influence 204 J. R. Nielsen et al. the overall results. Even if there were a potential effect, it would on analyses of variability in growth and maturity dynamics. In only affect the results and introduce noise concerning the addition to the work of Fraser et al. (2007), fishery landings sta- geographically (spawning area) related analyses because the main tistics do not indicate depth differences between sizes, and the data filling only concerned the geographical area disaggregation number of individuals found in deeper waters remains very level. low. Fishers have found no signs of emigration of the stock The conclusions are based on the assumption that there is no out of the area, and they have not noticed size- or age-specific significant difference or bias in NP catchability according to age patterns in occurrence according to depth either in the bank (for the 1+ group) or year in the surveys, i.e. the sampling of areas or along the Norwegian Trench (Flemming Christensen, each age group, especially the 1+ group, is representative for the a long-time NP fisher and former Chairman of the Danish stock. This assumption is assessed to be reasonable. Commercial Fishery Association, pers. comm.). Finally, papers First of all, age group 1 NP has a mean length of 11–15 cm in (Cook, 1997; Cotter, 2001) and reports (Beare et al., 2002; Q1–Q4 varying with sex, maturity, and region (Lambert et al., ICES, 2004) find similar trends in age-specific mortality using 2009), and the observed length range of age 1 in the full IBTS independent sources, i.e. different North Sea surveys and com- Q1 ALK time-series is 5–17 cm, of which only 0.6% are ,8 cm, mercial fishery data time-series. 1.6% are 8–9 cm, and 6.5% are 9–10 cm (DATRAS, www.ices. Consequently, age-, season-, and area-specific mortality pat- dk). Go¨tz (1997) published the only available selection parameters terns cannot be explained by survey coverage and catchability or

and ogive for NP in the IBTS GOV trawl, and she estimated an L50 by age-specific migration out of the area or vertical distribution Downloaded from of 8.0 cm, with a very narrow selection range, where 100% of the patterns by age. We have no objective information indicating NP are caught at length 9.2 cm, the length where North Sea NP are that larger NP (at least age 1+) are not representatively sampled fully selected by the IBTS survey gear. in the analysed IBTS surveys and that the constant catchability Second, there is no indication in the literature of lower catch- assumption is not valid. ability of the older age groups for the species covered by the IBTS survey, including NP (Cook, 1997; Cotter, 2001; Beare http://icesjms.oxfordjournals.org/ H02: there is no relationship between natural mortality et al., 2002; Anon., 2004). Cotter (2001), Cook (1997), and and the reproduction-specific life-history traits of NP, Anon. (2004) have indicated that survey catchability of 0-group such as sex, maturity, and growth, and thus, mortality gadoids, clupeoids, etc. may, in general, be relatively low due to is decoupled from spawning the mesh-size selection in the small-meshed IBTS survey trawls, This hypothesis is also rejected. The present study and Lambert but this is not the case for the 1+ group. Cook (1997) found et al. (2009) provide evidence that spawning mortality impacts that the small gadoid whiting has equal catchability for all ages the life-history traits and population dynamics of the NP stock. 1–6 in the IBTS survey. Furthermore, neither the surveys nor The ratio of mature individuals declines significantly from the commercial fleet have been able to find old NP (see below). before to after spawning, and only very few post-spawning NP Fraser et al. (2007) calculated the IBTS survey (GOV-trawl) catch- have ever been observed despite extensive surveying and fishing at DTU Library on February 21, 2012 ability for NP. The results showed a constant catchability for NP in in the North Sea. For the youngest age classes, the proportion of the length interval 12–20 cm, and the catchability was low for mature individuals is higher for males than for females, and small fish of 7.5 cm (the 0-group), which is similar to the total male mortality is higher. This is in accordance with Cooper length of the estimated L for NP for the GOV trawl (Go¨tz, 50 (1983), who found an increasing numerical dominance of NP 1997). Fraser et al. (2007) estimated lower IBTS catchability for females with age. Maturity and growth dynamics (Lambert et al., fish of lengths of 20–23 cm. However, the confidence limits 2009) strongly indicate greater mortality in the spawning areas for this estimate were high and substantially overlapped those and during the spawning season, as further discussed below. for the estimated catchability within the full size range 15– 23 cm, i.e. this result was not significant. NP at age 2 has mean lengths of 12–15 cm (immature fish) and 15–18 cm (mature), Geographical patterns and mortality dynamics whereas age 3 mean lengths are 14–16 and 18–20 cm for imma- pertaining to H02 and H03 ture and mature fish, respectively (Lambert et al., 2009). Geographical patterns and subarea-dependent mortality have not Therefore, the results of Fraser et al. (2007) did not indicate low been explored to the fullest extent in the present study using IBTS IBTS catchability of the age and length groups for which we esti- data because potential patterns therein might be flawed by mated high mortality, i.e. for ages 1–3. subarea-specific NP migrations within the North Sea and Third, several scientists have suggested that the depth distri- Skagerrak–Kattegat. Although such internal migration patterns bution of NP could increase with age (e.g. Poulsen, 1968; Raitt are not fully mapped, it is clear that Skagerrak–Kattegat is a and Mason, 1968; Albert, 1994, in the Norwegian Deep). nursery area and that NP migrate to the North Sea when maturing Sparholt et al. (2002a, b) analysed and discussed these potential (Ursin, 1963; Poulsen, 1968; Lambert et al., 2009). Geographically catchability changes with age in relation to depth-dependent dis- determined growth patterns of decreasing mean weight and length tribution and migration based on several sources, including with age in the spawning areas during the spawning season have IBTS data analyses and a literature review. They concluded been observed (Lambert et al., 2009). In addition, geographical that there was no evidence of vertical migration and associated maturity patterns have shown a significantly higher percentage age-specific migrations out of the NP population area and the of mature individuals in spawning areas RFA1 and RFA3, in IBTS area according to depth or topographical conditions, which there were significant decreases in the maturity ratio from which could explain the very less number of old NP in the Q1 to Q3, and where more than 90% of the spawners were catches in the North Sea and Skagerak–Kattegat. Furthermore, recorded in Q1 (present study and Figures 2 and 3 and Table 3 Lambert et al. (2009) demonstrated that there is no basis for in Lambert et al., 2009). This indicates that the larger, more dividing the stock into several smaller stock components based mature individuals disappear after spawning. It is also observed Do Norway pout (Trisopterus esmarkii) die from spawning stress? 205 that total mortality is significantly correlated with the percentage (saithe, haddock, and mackerel) in the IBTS Q3 survey. Both of of mature fish. Mortality cannot be directly calculated in the these studies are based on extrapolations of the 1991 “Year of spawning areas during and just after the spawning period, but the Stomach Sampling” diet compositions of predators. the results show that the yearly total mortality for both sexes is sig- However, strong predator–prey relationships do exist between nificantly positively correlated with the overall maturity ratio some commercially important North Sea stocks and NP. Adult assessed during the spawning season (Figure 3). The above whiting is an important predator of small NP (Jones, 1954; factors indicate a higher natural mortality associated with Daan and Welleman, 1998). In recent years, a significant part of spawning. the western mackerel stock has migrated to the North Sea, result- Based on stomach-content data analyses disaggregated to ICES ing in a potential higher predation mortality of small/young NP statistical square (area) and quarter of the year in the North Sea (particularly of the 0-group). Further, the North Sea saithe stock (1991), Rindorf et al. (2010) calculated biomass eaten and local has recently increased, leading to potentially higher predation predation mortality indices. They found that predated biomass mortality among larger NP (ICES, 2010). The stomach contents (and predation mortality) of NP by cod, whiting, haddock, and of the main predators should be analysed for the years beyond saithe was high in the second half of the year (Q4 and Q3) and 1991 (ICES, 2006, 2008; Kempf et al., 2009; Rindorf et al., 2010) low in the first half (Q2 and Q1). In Q1, the small NP biomass at the precise NP spawning time and place to determine whether eaten occurred in the most northern areas west of Orkney and NP are subject to increased predation when potentially weakened

south of Shetland. Based on Rindorf et al. (2010, Figures 2b and by spawning events. Downloaded from 5b), the areas of highest biomass predated and highest predation Although our analyses indicate density-dependent mortality mortality were not in the main spawning areas during the spawn- which can be associated with spawning and that available docu- ing season (Q1) that were identified by Lambert et al. (2009, e.g. mentation on predation cannot explain the observed increase in Figure 1). The latter study includes a review of previous studies Z at age, it is difficult to disentangle density-dependent mortality on NP spawning and egg/larvae distribution and identifies the and size-selective mortality. Size-selective mortality will usually main NP spawning areas to be in proximity to the 120-m isobaths result in greater mortality of the smallest (youngest) fish, but for http://icesjms.oxfordjournals.org/ in RFA1 and RFA3 near Viking Bank along the Norwegian NP, we observe greater mortality rates for the largest (oldest) Trench and along the Scottish east coast (and in RFA7) in Q1. fish, and that spawning is not only associated with age, but Consequently, predated biomass and predation mortality are low also with size. We find evidence of spawning mortality where in the main spawning areas and during the spawning season, indi- the fastest growing individuals mature faster and therefore cating that increased mortality cannot be explained by predation spawn and die faster, but there may be other reasons for mortality. such reversal size-selective mortality, e.g. density-dependence. Density-dependence probably does not influence mortality direct- H03: there is no density-dependence, intraspecific, ly, but rather indirectly as explained above, and can also be influ- or interspecific mortality in NP enced by size-selective mortality other than spawning mortality, at DTU Library on February 21, 2012 This hypothesis cannot be conclusively rejected. The density- so no rigorous conclusions can be made on the rejection of dependence, either intra- or interspecific, of NP mortality shows hypothesis H03. a distinct pattern. Mortality is significantly positively correlated with intraspecific population density. The NP population dyna- Conclusions and future studies mics seem, therefore, to be influenced by density-dependence, Our results indicate that a significant proportion of the NP stock which results in a lower growth rate and maturation when the most likely dies as a direct or indirect result of spawning. However, stock is at a relatively high level. Thus, bringing together the the variation in total mortality is high and cannot be exclusively varied information pertaining to NP mortality, it is likely that explained by this one life-history trait, i.e. other types of size- lower stock densities contribute to higher growth rates and selective mortality may also have an effect. In fisheries and ecosys- higher maturity ratios and, consequently, greater mortality rates, tem management, it is important to recognize the biological and which are most likely caused by spawning. Kempf et al. (2009) ecological contexts and mechanisms that lead this small, short- found no intraspecific relationship between NP SSB in the year lived gadoid to allocate so much energy to reproduction the first of birth and the IBTS age 1 recruitment index of the following time it spawns and to produce a high likelihood of death as a year, whereas the interannual variability in age 1 recruitment result of spawning stress or increased exposure to other mortality was found to be correlated with the Q2 sea surface temperature associated with spawning. Ursin’s (1963) studies on NP growth when taking predation impact into consideration. However, this have indicated that a likely cause of the observed growth and was not highly significant and included the removal of years char- energy allocation dynamics may be the mortality associated with acterized as outliers. spawning. Interspecific density-dependence and predation were not With respect to NP, future investigations should concentrate on significant factors based on the available data at the scale of our (i) intensified surveying of NP and its predators during and just study, but additional studies are necessary on more disaggregated after the spawning event at the spawning localities to follow mor- coverage and overlapping distribution and density patterns tality and predation patterns, (ii) precise maturity patterns and between NP and its main predators by age or size group, especially histological gonad analyses during the spawning season to follow during the spawning period. With regard to the overlap between the mortality patterns of NP that are closely associated with devel- NP and important predators in the North Sea, Rindorf et al. opment in the mature stages before and after spawning, (iii) histo- (2010) found low predated biomass and predation mortality in logical analyses of NP gonads and other organs in relation to the main spawning areas during the spawning season. Kempf potential senescence associated with spawning, and (iv) tank et al. (2009; Figure 10) found no strong correlation between the experiments on spawning NP. With these approaches, it will be spatial overlap of NP age 1 abundance and certain NP predators possible to evaluate some mortality mechanisms and to what 206 J. R. Nielsen et al. extent NP is weakened by the spawning event due to energy loss Cooper, A. 1983. The reproductive biology of poor cod, Trisopterus and increased vulnerability and exposure to fish and invertebrate minutus L., whiting, Merlangius merlangus L., and Norway pout predation, as well as senescence, sickness, and other residual Trisopterus esmarkii Nilsson, off the west coast of Scotland. mortality induced by spawning. Journal of Fish Biology, 22: 317–334. Cotter, A. J. R. 2001. Intercalibration of North Sea International Bottom Trawl Surveys by fitting year-class curves. ICES Journal Supplementary material of Marine Science, 58: 622–632. The following supplementary material is available at the ICESJMS Daan, N., and Welleman, H. 1998. Feeding ecology of the North Sea online version of the manuscript: bootstrap analysis and estima- fish with emphasis on the data base of the “Stomach Sampling tion of observation variability (CV) in the NP IBTS Q1 and Q3 Project 1991” for use in multispecies assessment. Netherlands Institute for Fisheries Research, Report C029/98. cpue data by age, and total mortality (Z) at age with confidence Fraser, H. M., Greenstreet, S. P. R., and Piet, G. J. 2007. Taking account limits from a SURBAR full population dynamic model assessment of catchability in groundfish survey trawls: implications for esti- in 2011. mating demersal fish biomass. ICES Journal of Marine Science, 64: 1800–1819. Acknowledgements Gislason, H., Daan, N., Rice, J. C., and Pope, J. G. 2010. Size, growth, temperature and the natural mortality of marine fish. Fish and We would like to thank all the international fisheries research insti- Fisheries, 11: 149–158. tutes for participating in sampling and compiling the ICES IBTS Downloaded from Golubev, A. 2009. How could the Gompertz–Makeham law evolve. data and for letting us use the data in the present context. Journal of Theoretical Biology, 258: 1–17. Thanks also to ICES for providing the raw data. We would also Go ¨tz, S. 1997. Aspekte Zur Nahrungso¨kologie demersaler Nordseefische. like to thank Hans Lassen, ICES Secretariat, for valuable help Institut Fu¨r Hydrobiologie und Fishereiwissenschaft, Diplomarbeeit, with evaluating IBTS cpue variability, Coby Needle, MARLAB Universita¨t Hamburg, Germany. 93 + VI pp. (Report under the Scotland for an updated NP SURBAR Assessment, and Henrik EU AIR3-CT94-2410 Contract and Project). Gislason, DTU Aqua, for inspiring discussions on fish mortality. Hoenig, J. M., and Heisey, D. M. 1987. Use of a log-linear model with http://icesjms.oxfordjournals.org/ EM algorithm to correct estimates of stock composition and to convert length to age. Transactions of the American Fisheries References Society, 16: 232–243. Albert, O. T. 1994. Biology and ecology of Norway pout (Trisopterus ICES. 2004. Report of the Working Group on the Assessment of esmarkii Nilsson, 1855) in the Norwegian Deep. ICES Journal of Demersal Stocks in the North Sea and Skagerrak (WGNSSK), Marine Science, 51: 45–61. Benchmark Assessment. ICES Document CM 2004/ACFM: 07. Anon. 2001. DATRAS, DAtabase TRAwl Surveys, Final Report. 70 pp. ICES. 2006. Report of the Study Group on Multispecies Assessments in www.ices.dk. the North Sea (SGMSNS), 20–25 February 2006, ICES Anon. 2004. Manual for the International Bottom Trawl Surveys Headquarters, Copenhagen. ICES Document CM 2006/RMC: 02.

(IBTS), Revision VII. International Bottom Trawl Survey 75 pp. at DTU Library on February 21, 2012 Working Group, ICES. 51 pp. http://www.ices.dk/datacentre/ ICES. 2007a. Report of the Working Group on the Assessment of datras/trawldetails.asp. Demersal Stocks in the North Sea and Skagerrak (WGNSSK). Bailey, R. S., and Kunzlik, P. A. 1984. Variation in growth and mortal- ICES Document CM 2007/ACFM: 07. ity rates of Norway pout Trisopterus esmarkii (Nilsson). ICES ICES. 2007b. Report of the Ad Hoc Group on Real Time Management Document CM 1984/G: 70. and Harvest Control Rules for Norway Pout in the North Sea and Bass, T. M., Weinkove, D., Houthoofd, K., Gems, D., and Partridge, L. Skagerrak (ICES AGNOP). ICES Document CM 2007/ACFM: 39. 2007. Effects of resveratrol on lifespan in Drosophila melanogaster ICES. 2007c. Report of the Ad Hoc Group on Sandeel and Norway and Caenorhabditis elegans. Mechanisms of Ageing and Pout (ICES AGSANNOP). ICES Document CM 2007/ACFM: 40. Development, 128: 546–552. ICES. 2007d. Confidence Limits Estimation of Abundance Indices Baur, J. A., Pearson, K. J., Price, N. L., Jamieson, H. A., Lerin, C., Kalra, from Bottom Trawl Survey Data—Implementation in DATRAS. A., Prabhu, V. V., et al. 2006. Resveratrol improves health and sur- ICES Report to EU Commission on Survey Variance, ICES vival of mice on a high-calorie diet. Nature, 444: 337–342. Headquarters, February 2007. www.ices.dk. Beare, D., Castro, J., Cotter, J., van Keeken, O., Kell, L., Laurec, A., Mahe´, J. C., et al. 2002. Evaluation of research surveys in relation ICES. 2008. Report of the Working Group on Multispecies Assessment to management advice. Final Report, EU EVARES Project Methods (WGSAM), 6–10 October 2008, ICES Headquarters, (FISH/2001/02 – Lot1), EU Commission. 307 pp. Copenhagen. ICES Document CM 2008/RMC: 06. 113 pp. Beverton, R. J. H. 1963. Maturation, growth, and mortality of clupeoid ICES. 2009. Report of the ICES Advisory Committee, 2009. ICES and engraulid stocks in relation to fishing. Rapport et Advice, 1–11. 1420 pp. Proce`s-Verbaux des Re´unions du Conseil Permanent ICES. 2010. Report of the Working Group on the Assessment of International pour l’Exploration de la Mer, 154: 44–67. Demersal Stocks in the North Sea and Skagerrak (WGNSSK). Buston, P. M., and Garcia, M. B. 2007. An extraordinary life span es- ICES Document CM 2010/ACOM: 13. timate for the clown anemonefish Amphiprion percula. Journal of Jakobsen, T., and Ajiad, A. 1999. Management implication of sexual Fish Biology, 70: 1710–1719. differences in maturation and spawning mortality of Northeast Caputo, V., Candi, G., Arneri, E., Mesa, M. L., Cinti, C., Provinciali, Arctic cod. Journal of Northwest Atlantic Fishery Science, 25: M., Cerioni, P. N., et al. 2002. Short lifespan and apoptosis in 125–131. Aphia minuta. Journal of Fish Biology, 60: 775–779. Jones, R. 1954. The food of the whiting and a comparison with that of Cook, R. M. 1997. Stock trends in six North Sea stocks as revealed by the haddock. Marine Research, 2: 1–34. an analysis of research vessel surveys. ICES Journal of Marine Kempf, A., Floeter, J., and Temming, A. 2009. Recruitment of North Science, 54: 924–933. Sea cod (Gadus morhua) and Norway pout (Trisopterus esmarkii) Cook, R. M. 2004. Estimation of the age-specific rate of natural mor- between 1992 and 2006: the interplay between climate influence tality for Shetland sandeels. ICES Journal of Marine Science, 61: and predation. Canadian Journal of Fisheries and Aquatic 159–169. Sciences, 66: 633–648. Do Norway pout (Trisopterus esmarkii) die from spawning stress? 207

Kishi, S., Uchiyama, J., Baughman, A. M., Goto, T., Lin, M. C., and Sparholt, H. 1994. Fish species interactions in the Baltic. Dana, 10: Tsai, S. B. 2003. The zebrafish as a vertebrate model of functional 131–162. aging and very gradual senescence. Experimental Gerontology, Sparholt, H., Larsen, L. I., and Nielsen, J. R. 2002a. Verification of 38: 777–786. multispecies interactions in the North Sea by trawl survey data Lambert, G., Nielsen, J. R., Larsen, L. I., and Sparholt, H. 2009. on Norway pout (Trisopterus esmarkii). ICES Journal of Marine Maturity and growth population dynamics of Norway pout Science, 59: 1270–1275. (Trisopterus esmarkii) in the North Sea, Skagerrak and Kattegat. Sparholt, H., Larsen, L. I., and Nielsen, J. R. 2002b. Non-predation ICES Journal of Marine Science, 66: 1899–1914. natural mortality of Norway pout (Trisopterus esmarkii) in the Partridge, L. 2010. The new biology of ageing. Philosophical North Sea. ICES Journal of Marine Science, 59: 1276–1284. Transactions of the Royal Society of London, Series B, 365: Sparre, P., and Venema, S. C. 1989. Introduction to tropical fish stock 147–154. assessment. Part 1—manual. FAO Fisheries Technical Paper, 306/ Poulsen, E. M. 1968. Norway pout: stock movements in the Skagerrak 1. 337 pp. and the north-eastern North Sea. Rapports et Proce`s-Verbaux des Re´unions du Conseil International pour l’Exploration de la Mer, Terzibasi, E., Valenzano, D. R., and Cellerino, A. 2007. The short-lived 158: 80–85. fish Nothobranchius furzeri as a new model system ageing studies. Raitt, D. F. S., and Mason, J. 1968. The distribution of Norway pout in Experimental Gerontology, 42: 81–89. the North Sea and adjacent waters. Marine Research, 4: 1–19. Uriarte, A., Ibaibarriaga, P. L., Abaunza, P., Pawlosky, L., Masse´, J., Reznick, D., Bryant, M., and Holmes, D. 2006. The evolution of sen- Petitgas, P., Santos, M., et al. 2010. Assessing natural mortality of

escence and post-reproductive lifespan in guppies (Poecilia reticu- anchovy from surveys population and biomass estimates. ICES Downloaded from late). PLoS Biology, 4(1, e7): 136–143. Document CM 2010/C: 12. 25 pp. Ricker, W. E. 1975. Computation and interpretation of biological sta- Ursin, E. 1963. On the seasonal variation of growth rate and tistics of fish populations. Bulletin of the Fisheries Research Board growth parameters in Norway pout (Gadus esmarkii) in the of Canada, 191. 382 pp. Skagerrak. Meddelelser fra Danmarks Fiskeri-og Rindorf, A., Andersen, N. G., and Vinther, M. 2010. Spatial differences Havundersogelser, 4: 17–29.

in natural mortality of North Sea gadoids. ICES Document CM Woodhead, A. D. 1998. Aging, the fishy side: an appreciation of Alex http://icesjms.oxfordjournals.org/ 2010/C: 18. Comfort’s studies. Experimental Gerontology, 33: 39–51.

Handling editor: Emory Anderson at DTU Library on February 21, 2012